INTRODUCTION: The outbreak-detection performance of a syndromic surveillance system can be measured in terms of its ability to detect signal (i.e., disease outbreak) against background noise (i.e., normally varying baseline disease in the region). Such benchmarking requires training and the use of validation data sets. Because only a limited number of persons have been infected with agents of biologic terrorism, data are generally unavailable, and simulation is necessary. An approach for evaluation of outbreak-detection algorithms was developed that uses semisynthetic data sets to provide real background (which effectively becomes the noise in the signal-to-noise problem) with artificially injected signal. The injected signal is defined by a controlled feature set of variable parameters, including size, shape, and duration. OBJECTIVES: This report defines a flexible approach to evaluating public health surveillance systems for early detection of outbreaks and provides examples of its use. METHODS: The stages of outbreak detection are described, followed by the procedure for creating data sets for benchmarking performance. Approaches to setting parameters for simulated outbreaks by using controlled feature sets are detailed, and metrics for detection performance are proposed. Finally, a series of experiments using semisynthetic data sets with artificially introduced outbreaks defined with controlled feature sets is reviewed. RESULTS: These experiments indicate the flexibility of controlled feature set simulation for evaluating outbreak-detection sensitivity and specificity, optimizing attributes of detection algorithms (e.g., temporal windows), choosing approaches to syndrome groupings, and determining best strategies for integrating data from multiple sources. CONCLUSIONS: The use of semisynthetic data sets containing authentic baseline and simulated outbreaks defined by a controlled feature set provides a valuable means for benchmarking the detection performance of syndromic surveillance systems.

BACKGROUND: The Newborns' and Mothers' Health Protection Act of 1996 prohibits payers from restricting "benefits for any hospital length of stay in connection with childbirth for the mother or newborn child, following a normal vaginal delivery, to less than 48 hours." The law recognizes the basic right of women and physicians to make decisions about aptness of discharge timing. OBJECTIVE: To provide data as a basis for decisions about aptness of discharge timing by studying the effect of voluntary, moderate reductions in length of postpartum hospital stay on an array of maternal and infant health outcomes. DESIGN: A prospective cohort study. Patients were surveyed by telephone at 3 and 8 weeks postpartum. SETTING: A teaching hospital where 38% of the patients are in a managed care health plan with a noncompulsory reduced stay program offering enhanced prepartum and postpartum services, including home visits. PATIENTS: Consecutive mothers discharged after vaginal delivery during a 3-month period. MAIN OUTCOME MEASURES: The outcomes were health services use within 21 days, breast-feeding, depression, sense of competence, and satisfaction with care. Multivariate analyses adjusted for sociodemographic factors, payer status, services, and social support. RESULTS: Of 1364 eligible patients, 1200 (88%) were surveyed at 3 weeks; of these 1200, 1015 (85%) were resurveyed at 8 weeks. The mean length of stay was 41.9 hours (SD, 12.2 hours). Of patients going home in 30 hours or less, 60.8% belonged to a managed care health plan. The length of stay was not related to the outcomes, except that women hospitalized shorter than 48 hours had more emergency department visits than those staying 40 to 48 hours (adjusted odds ratio, 5.78; 95% confidence interval, 1.19-28.05). CONCLUSIONS: When adequate postpartum outpatient care is accessible, a moderately shorter length of postpartum stay after an uncomplicated vaginal delivery had no adverse effect on an array of outcomes. Researchers and policy makers should seek to better define the content of postpartum services necessary for achieving optimal outcomes for women and newborns; funding should be available to provide such services, regardless of the setting in which they are provided.

BACKGROUND: The Centers for Disease Control and Prevention is incorporating laboratory data into real-time surveillance systems. When normal patterns of laboratory test orders and results are modeled, aberrations can be detected. Because many test orders are available electronically well before results, atypical patterns of test ordering may signal outbreaks. OBJECTIVES: The authors sought to characterize baseline patterns in the ordering and early results of lumbar punctures, motivated by the possibility of using these data for real-time surveillance for early detection of meningitis or encephalitis outbreaks. METHODS: Retrospective cohorts of pediatric emergency department patients at a single hospital (1993-2003) and from the National Hospital and Ambulatory Medical Care Survey (1992-2000) were used for analysis. RESULTS: Test ordering exhibits seasonal patterns, with monthly peaks in January and August (p 0.0001). For the hospital cohort, the rate of cerebrospinal fluid pleocytosis exhibits seasonal patterns (p 0.0001), with a peak from August to October. This is strongly associated with the rate and pattern of clinical neurologic disease (p 0.0001). A long-term secular decline in daily test ordering is evident, dropping from 5.3 to 2.9 in the hospital sample, and from 371.8 to 185.3 in the national sample (p 0.001). The long-term rate of pleocytosis has declined (p 0.0001), though the yield of testing for pleocytosis has improved (p = 0.0104). CONCLUSIONS: Laboratory test patterns correspond with those of clinical disease and are a promising source of surveillance data. Using such data for real-time monitoring requires specific adjustments for patient age, periodicities, and secular trends.

OBJECTIVE: To determine whether longitudinal data in patients' historical records, commonly available in electronic health record systems, can be used to predict a patient's future risk of receiving a diagnosis of domestic abuse. DESIGN: Bayesian models, known as intelligent histories, used to predict a patient's risk of receiving a future diagnosis of abuse, based on the patient's diagnostic history. Retrospective evaluation of the model's predictions using an independent testing set. SETTING: A state-wide claims database covering six years of inpatient admissions to hospital, admissions for observation, and encounters in emergency departments. Population All patients aged over 18 who had at least four years between their earliest and latest visits recorded in the database (561,216 patients). MAIN OUTCOME MEASURES: Timeliness of detection, sensitivity, specificity, positive predictive values, and area under the ROC curve. RESULTS: 1.04% (5829) of the patients met the narrow case definition for abuse, while 3.44% (19,303) met the broader case definition for abuse. The model achieved sensitive, specific (area under the ROC curve of 0.88), and early (10-30 months in advance, on average) prediction of patients' future risk of receiving a diagnosis of abuse. Analysis of model parameters showed important differences between sexes in the risks associated with certain diagnoses. CONCLUSIONS: Commonly available longitudinal diagnostic data can be useful for predicting a patient's future risk of receiving a diagnosis of abuse. This modelling approach could serve as the basis for an early warning system to help doctors identify high risk patients for further screening.

OBJECTIVE: Broadly, to create a bidirectional communication link between public health surveillance and clinical practice. Specifically, to measure the impact of integrating public health surveillance data into an existing clinical prediction rule. We incorporate data about recent local trends in meningitis epidemiology into a prediction model differentiating aseptic from bacterial meningitis. DESIGN AND MEASUREMENTS: Retrospective analysis of a cohort of all 696 children with meningitis admitted to a large urban pediatric hospital from 1992 to 2000. We modified a published bacterial meningitis score by adding a new epidemiological context adjustor variable. We examined 540 possible rules for this adjustor, varying both the number of aseptic meningitis cases that needed to be seen, and the recent time window in which they were seen. We performed sensitivity analyses with each of 540 possibilities in order to identify the optimal rule--namely, the one that included the most cases of aseptic meningitis without missing additional cases of bacterial meningitis, as compared with the published prediction model. We used bootstrap methods to validate this new score. RESULTS: The optimal rule was found to be: "at least four cases of aseptic meningitis in the previous 10 days." The epidemiological context adjustor based on surveillance of recent cases of meningitis allowed the correct identification of an additional 47 cases (7%) of aseptic meningitis without missing any additional cases of bacterial meningitis. The epidemiological context adjustor was validated, showing significance in 84% of 1,000 bootstrap samples. CONCLUSION: Epidemiological contextual information can improve the performance of a clinical prediction rule. We provide a methodological framework for leveraging regional surveillance data to improve medical decision-making.

The BiliLIGHT system is a World Wide Web (Web) based system that integrates an interactive clinical practice guideline with real-time patient-data retrieval from remote heterogeneous data sources to help clinicians manage newborn jaundice at the point of care in three clinical settings. We briefly describe the system, how actual information exchange of medical data across institutional boundaries was achieved, and how the data were connected to a CPG. In particular, we examine the requirements for patient identification, exchange protocols, authentication, and a standard vocabulary.

OBJECTIVE: The purpose of this study was to evaluate the effects of an early postpartum discharge program and a subsequent legislative mandate for 48 hours of hospital coverage on incidence of newborn jaundice and feeding problems. We tested the hypothesis that heightened postdischarge ascertainment (rather than short stays) is responsible for apparent increases in these outcomes. METHODS: Interrupted time series analysis was conducted on retrospective data from the automated medical records of a large Massachusetts health maintenance organization (HMO). A population of 20,366 mother-infant pairs with normal vaginal deliveries between October 1990 and March 1998 was identified. The interventions included a new HMO protocol in 1994 of 1 hospital overnight after delivery, plus a nurse home visit, then the Massachusetts' 1996 minimum coverage law. Postpartum length of stay, clinical evaluation on day 3 or 4 of life, health center visits up to day 21, health center diagnoses of jaundice or feeding problems, bilirubin testing and test severity, rehospitalizations, and emergency department visits were measured. RESULTS: Postpartum stays 2 nights rose from 28% of newborns before implementation of the program to 70% immediately after implementation. Later, this rate fell from 66% before the mandate to 21% just after the law went into effect. Day 3 or 4 evaluation rose from 24.5% to 64% after the program, then dropped somewhat to 53% after the mandate. Controlling for longer-term trends in health center visits, implementation of the early discharge program was associated with approximately 1 extra visit for every 4 newborns within the first 21 days of life. The state mandate did not affect health center visit rates. Jaundice diagnoses were flat at 8% of newborns during the baseline, then rose to a constant 11% throughout the program and postmandate periods. Bilirubin testing of newborns also rose by 3.4 percentage points at the time of program implementation, and the proportion of tested newborns with results calling for at least consideration of phototherapy rose by 6 percentage points. Phototherapy use rose from a flat 1.8% to 2.4% of newborns after program implementation. Feeding problem diagnoses more than doubled at the time of program implementation and remained elevated after the mandate. Rehospitalizations overall and specifically for jaundice were constant over time, whereas more rare emergency department visits for jaundice dropped from 0.3% of newborns to 0 on program implementation. CONCLUSIONS: Sudden increases in jaundice-related measures and identification of infant feeding problems were not associated with changes in length of stay in this setting. Instead, these increases seem to be the result of more frequent evaluation of newborns during the critical day 3 to 4 period and may also have been elevated by a new climate of concern about neonatal vulnerability. "Ascertainment bias" may have confounded findings in previous reports that raised concerns about the safety of early discharge.

Open sharing of clinical trial data has been proposed as a way to address the gap between the production of clinical evidence and the decision-making of physicians. A similar gap was addressed in the software industry by their open-source software movement. Here, we examine how the social and technical principles of the movement can guide the growth of an open-source clinical trial community.